Wassim Mohsni

Implementing a microservice architecture-based project that leverages the power of Google cloud OCR to analyze documents (invoices, bank statements, etc...), along with a complete toolset allowing accountants to efficiently organize and execute their work.

Designing, implementing and maintaining a real-time deep learning medical imagery analytics application using Nvidia Deepstream SDK (C++), GTK+ and Nvidia Jetson. •Designed advanced dynamic Gstreamer pipelines with continuously and seamlessly changing elements at runtime. •Achieved high QoL and better user satisfaction successfully attracting new clients.

- Implemented and maintained an End-to-End deep learning pipeline for the National Traffic census of 2022/2023 in cooperation with the Tunisian Ministry of equipment housing and infrastructure. It is deployed in more than 25 geographical regions all over the Tunisian country (edge devices) and on a private dedicated server. - Designed, Developed, and maintained a web platform (backend, frontend, and database management) for the National Traffic Census Deep learning pipeline, the platform allows for edge device monitoring and configuration, and data visualization, and video upload for cloud inference.

Implemented a Deep learning-based near real-time Tunisian license plate recognition system (ANPR)

Automated Digitization of documents based on Tesseract OCR

Development of a back office for user-related reservation management